Nanoscale remodeling of synaptic glutamate receptors and calcium channels in temporal lobe epilepsy

Abstract

Temporal lobe epilepsy (TLE) is the most common form of focal epilepsy and remains drug-resistant in approximately one-third of patients, for whom surgical resection of the epileptogenic zone is currently the most effective treatment option. One strategy to improve therapeutic outcomes involves identifying the molecular mechanisms and key regulators underlying synaptic function. Although altered glutamatergic neurotransmission is considered a major contributor to epileptic hyperexcitability, the nanoscale organization of key synaptic proteins in the human epileptic cortex is still not well characterized. In particular, it remains unclear whether pre- and postsynaptic proteins undergo synapse-type-specific remodeling in TLE. Here, we used high-resolution freeze-fracture replica immunolabeling to map the nanoscale organization of postsynaptic AMPA receptors (AMPARs) and presynaptic P/Q-type voltage-gated calcium channels (CaV2.1) in glutamatergic synapses of cortical layers II/III of surgically resected human tissue. We examined perilesional, cytologically unremarkable, non-epileptogenic cortical tissue obtained from patients undergoing tumor resection (52-63 years; two female and three male patients) and compared these data with cortical samples from patients with refractory TLE (22-53 years; two female and four male patients). To assess whether pathological remodeling differs across synapse populations, we analyzed both excitatory dendritic spine and shaft synapses. We found a robust, highly consistent increase in AMPAR density at both spine and shaft synapses in TLE patients indicating enhanced postsynaptic excitatory signaling. However, AMPAR spatial reorganization differed between synapse types. Monte Carlo simulations demonstrated that spine synapses largely preserved a random receptor distribution without a significant clustering pattern, whereas shaft synapses exhibited a redistribution from clustering toward random receptor arrangements together with reduced center-preference within the postsynaptic density. In contrast, presynaptic CaV2.1 remodeling was more variable and predominantly associated with excitatory spine synapses. TLE spine synapses exhibited an overall increase in CaV2.1 density together with significantly larger and more numerous CaV2.1 nanoclusters within the active zone. This difference was particularly pronounced in male patients. In comparison, shaft synapses displayed only minor presynaptic alterations. CaV2.1 channels remained highly clustered under both control and epileptic conditions, indicating preservation of core active-zone organizational features despite disease-associated nanoscale remodeling. Together, these findings provide direct ultrastructural support that pathological plasticity in human TLE involves an overall increase in excitatory postsynaptic responsiveness and a target cell-specific increase of excitatory release probability at cortical synapses.